Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
408085 | Neurocomputing | 2011 | 5 Pages |
Usually the functioning of a transcription factor to regulate a target gene requires a modifying enzyme. The triplet of transcription factor, mediating enzyme and target gene, MFG triplet for short, is like a switchboard that transmits signals from signaling network to gene regulatory network. Using PTM-Switchboard database, we systematically studied the MFG triplets in yeast and constructed a predictor to identify MFG triplets based on functional profiles of each element in the triplet. The sensitivity (Sn) and specificity (Sp) of Nearest Neighbor Algorithm (NNA) were 75.106% and 99.585%, respectively. Our method had higher sensitivity than Bayesian network (NaiveBayes), decision tree (J48), Random Forest (RandomForest) and k-nearest neighbor (KNN, K=3). The established predictor may be used to identify MFG triplet, which can link up cellular signaling transduction with transcriptional regulation, two major subjects in biological research.